Discussion of Tessitura
by Guest Author Beatrice

The only way to really know the tessitura of a piece of music or an entire operatic role is to go through the music and take note of where most of the notes lie. Knowing the range, or the highest and lowest notes, is not really helpful, as those notes can be altered if needed and the part still retain its correctness in interpretation.

There are many singers who have learned roles only to not sing them, knowing that as they learned them, even though the notes were well within their ranges, the vocal centre, or the basic tessitura of the role was not comfortable.

The great Mezzo-Soprano, Stignani, could easily sing Norma based on the notes of the score, for she had a very large full high C and it was quite effortless. But her voice sat much lower making such roles as Ulrica (a definite contralto part) equally as easy to sing. At the piano, Norma was easy for her, but she knew the difficulties that a singer faces when with full orchestra and chorus. The strength needed is quite a different thing.

Most students make the mistake of looking at the arias, or the famous arias, and seeing if they can sing them. But the real meat of the role is seldom the famous arias, beautiful as they are. The meat of the role is in the ensembles, the duets, the trios, and the recitatives. A role can take on a very different characteristic when those things are taken into account.

A role can even change when recitatives are added by other composers. The original French version of Cherubini’s Medee has spoken dialogue not recitatives. Medee is not a high role, and it progresses in intensity as the opera progresses. When Franz Lachner, a German organist, composed recitatives in 1855 (originally in German, and originally all the spoken dialogue set to music; later shortened somewhat and translated into Italian, the version most known today), the entire flavour of the opera changed. The flavour of a French Opera-comique of 1797 suddenly took on the heaviness of a German Romantic opera of the mid-1800s. The dramatic recitatives of the “new version” completely altered the tessitura and style of the opera. And in many respects actually weakened the drama considerably. Medea suddenly becomes a heavy dramatic soprano role, darker in nature, and with less development in the character (the original very carefully brings out Medee’s personality only slowly revealing her true nature). The demands on the lower register remained, but were interspersed with hectic screeching in the upper range. The original intensity of those higher notes, being saved until Medee’s actual real nature is revealed is lost. And the tessitura becomes very awkward. Aside from a very few, most singers who have song the role really cannot do it justice in this version.

Sometimes we have roles that the composer classified as Soprano or Mezzo-Soprano, but if the range is used as a measure of who should sing which, we see that both voices sing exactly the same notes. Such a case is Verdi’s “Oberto.” In this opera the soprano lead, Leonora, is really a very low role. She sings much of her music well within the staff. In ensembles she sings higher lines, as one would associate with a soprano, but even then, they are not all that high. She is required right off the bat to sing many resounding low As. Many of her arias begin on Middle C. Going by the range of notes, a singer would never know just how strong the lower parts of the soprano’s voice must be. If the soprano who sings this role doesn’t have powerful low notes, she will be ineffective.

Then, there is the Mezzo role of Cuniza. She covers the same notes as Leonora, but often doesn’t even go as low. Her first aria is not all that low, and has a few high notes, but it is concentrated more at the lower end of the staff.

Leonora descends lower in her arias, but they do sit in the staff a note or maybe a tone and a half higher than Cuniza’s do. The type of decorations are the same for both roles. Since based on arias both roles appear they can be sung by the same voice, what makes the two different (other than the composer requesting different voices)? It is again the ensembles, duets, and recitatives. For the most part, Cuniza’s recitatives are a touch lower than Leonora’s, but only a touch. It is in the ensemble pieces that we see the difference. Here Leonora will ascend the scale and have lines that sit above the staff. Her voice is to be heard above the others. Cuniza sits very much within the staff. Her contribution is necessary, and her voice cannot and should not become “silent” just because she is lower in sound (which often happens when singers are in lower ranges). While a darker soprano is needed for Leonora (no bright chirpers here), Cuniza is definitely the property of a mezzo or even a contralto. For her to be telling in an ensemble, it is a must she have a strong lower middle voice.

I think we can learn a great deal more about tessitura while dealing with a single role and its two main arias. The role being Urbain from Les Huguenots.

Before we begin, though, we must comment on the scores that are currently available. We must understand that many of them are not representative of what the composer’s final thoughts were, nor did many composers have a “final version” of many of their operas. Most of what we have are common performing scores. Sometimes the composer oversaw what we have printed, sometimes he did not. Sometimes he authorized cuts which didn’t get to the printer on time, other times, what the printer did was something entirely their own doing. Tradition has also played a huge factor in what we have printed.

Until we have more critical editions (which are being prepared for more and more operas of various composers) we cannot be totally sure of what was intended, and often get quite a wrong idea of the needs of a role. Sometimes, with the attempt to “keep all the music the composer wrote” for a given role, we end up with total confusion, for the role was altered depending on the singers, but usually in the editions we have, the alteration is not considered.

The result is total confusion regarding what is the real tessitura of the role.

We also have a great deal of trouble now days judging singers of the past basing what we think on the printed scores we have. This is not a good thing, and as is usually the case, the conclusions are quite inaccurate.

Thankfully to many musicologists, like Phillip Gosset, we are beginning to see what the composer intended, not only in what he originally set to paper, but often with all the variations he created for various singers who were interpreting the role. Some roles we find totally rewritten for a completely different voice range (Giulietta in I Capuletti e i Montecchi; originally a soprano, but altered to a contralto, while Romeo stays a mezzo-soprano). Also, because of much research, we are also finding the actual embellishments used by many singers as they sang the various roles. Often seeing just how much they transposed what they sang.

Before discussing the role of Urbain, I feel we need to discuss for a moment many of the great singers for whom many of the various roles were written, and how what we see printed is NOT what they actually sang.

Pasta was a great soprano for whom Bellini wrote both Norma and La Sonnambula, yet no two roles could be further apart in nature, or in ability, or in tessitura. Although Pasta had a respectable range which allowed her to sing a slightly sharpened D at times, mostly her highest note with comfort, was lucky to be a C. She could descend to a very resonant A below middle C with ease. However, most critics of her day did not see her as a soprano, but really a mezzo soprano, as her voice was dark, not very flexible, and had many flaws.

When you compare NORMA with LA SONNAMBULA and with the other roles she was famous for (Medea in Corinto by Mayr as an example), we see that few role she sang really extended to even a high C, but were well placed within the staff only occasionally venturing above it or below it. Even the version of La Sonnambula we have places the role of Amina more within the staff than above for most parts of it. That is until we get to the end, then suddenly, with “Ah! non giunge” we are all over the map and rising to high E flat. Certainly what we have is impressive. But Pasta had many issues with her voice: she had pitch problems in the passaggio, and was always flat on the notes E, F, and F sharp. Even her students remarked how they would learn music and when singing those notes, she would always chastize them for singing too sharp, but when they went to the piano, they were in tune. She could not sing a fluid scale straight up from a very low note to the highest note in the scale, but had to have it broken down into groups so as to allow her voice to alter itself as it ascended.

We all know that Bellini transposed the great Casta Diva from Norma down to F from his original G for Pasta, and the reason was in G too many of her always flat notes would occur, so he lowered it so it would be easier for her to stay in tune while she sang.

As we look at the critical editions of these scores we discover that Bellini wrote NO high E flats for Pasts, nor any high D flats either. Most of her music is written a tone lower than we have it currently published, and the great flights of fancy we have in “Ah! non giunge” simply do not exist. The original version is very simple, and quite in keeping with the pastoral nature of the work. The tenor’s music is usually written a tone higher than we have currently published, but Rubini, the creator of the tenor role often sang the music as much as a fourth higher than written taking him often to high E flat. But he did that only in contemplative moments, not usually in the heavily dramatic moments. In those moments he sang what was written. And he cared less what he sang in ensembles, as he often didn’t sing his music at all during such moments. Pasta and Rubini duets present us with various key differences (so both singers could be happy with their roles) and even various changes to the coloratura.

When placed back into its original keys, La Sonnambula is not all that different from Norma in tessitura. We can see the possibility that the same singer did create both roles. Because of our luck, really, of having many great singers of today with the ability to sing both roles as published, we have made erroneous conclusions as to how Pasta sounded. Callas and Sutherland were the best singers who did both roles, and each brought with them some wonderful singing. But neither was a Pasta.

Another singer we often associate with Amina in La Sonnambula is Maria Malibran. She is not a soprano, and was never designated such by any critic of her day. She was a contralto who forced her voice to acquire some of the high notes used by sopranos of that day, even to a high E. Her voice was never disciplined, was often unreliable, and seldom at her command. Yet, when she sang, she sang. She was Bellini’s favourite Amina. But did she sing what we have published? Not at all. Though she could hit a High E, and did in some operas she sang, she did NOT sing that note in La Sonnambula.

For the majority of the role, she lowered the music by a third. For some of the great coloratura numbers (Come per me sereno….sovra il sen la man) she lowered the music a fourth. Her added embellishments were extraordinary and quite exciting descending to F below middle C and ascending to a B. Her reading of “Ah! Non credea” is in G minor and again, her embellishments (which are almost too extreme for the aria’s simplicity, but still very tasteful and effective) are placed more at the lower end of the middle range. Patti recorded the aria in the “Malibran key” when she recorded it in later life. Her own embellishments, though modest, attesting to some of what singers did at that time. Lind also sang this role, but in the keys we are familiar with today, and again, her embellishments are more extensive than we hear in any recording, but well within the mood of the aria.

As for the great show piece at the end, Ah! Non giunge,” well, she (Malibran) lowered that to G major, and sang it with a wonder of embellishments, but respected at the same time the simplicity of the original. She would touch a high B a few times, but also descend to the low G below middle C and made a two octave jump to G above the staff. When taken as a whole, she placed the role definitely within the range of a mezzo-soprano, with occasional touching of notes slightly outside the comfort zone of that range.

The great contralto, Marietta Alboni, also sang La Sonnambula, and used the Malibran keys when she sang it, but her embellishments were far more modest, but extraordinarily beautiful.

So why are we using a score that is not correct? Because that is what ultimately was published in the days following the opera becoming quite popular. Most of the singers who sang the version we have today were what were called “the Nightingales.” They were noted as being quite high sopranos, and did show off their abilities. What we have in the score is a reading actually given of some of what Persiani sang in performance (though her actual embellishments were even more complicated, far outdoing anything we hear with Callas or even Sutherland). It stuck, and singers went from there to much more complicated embellishments.

What we have really is NOT a score reflecting the intentions of the composer, but a score that reflects the performing practices of the day. By so doing, we have a score that would really have been outside the abilities of the singer who created it. Pasta never changed her embellishments once she found those that reflected her desires. And from what has been recorded, she seldom embellished all that much. The flights of fancy we see in this final aria are not in keeping with even what she would have done.

Now days, there are a number of recordings out there that claim to use the Critical Edition of this score, and the results are quite beautiful, though NONE of them use a voice that would be like that of a Pasta or a Malibran (not even Bartoli with her delicate mezzo; neither of these two singers were delicate singers, but powerful singing actresses). Interestingly, most of the public don’t like this reading of the score because it is shorn of all its impressive embellishments, and they find the “Ah! Non giunge” a total let down with few embellishments and no high notes to speak of. Most singers will sing the written version in the first singing of the aria, or the first repeat, and then sing something like what we have printed (with embellishments) for the second repeat. A choice to please the public, who is used to a very decorated version of the opera. It would be nice to not only use the Critical Edition, but use the embellishments Pasta used, or those that Malibran used, and see how that would impress the public. It may impress with fireworks, but I think it would still disappoint because there would not be any high notes, and no high note endings we are so used to having.

We must also consider a moment how voices were trained back then. Even though many singers had incredible high notes, they were not the powerful notes we are familiar with today. They were penetrating, and they carried extremely well, but the volume and loudness we are accustomed to was seen as crude and uncouth. All voices, even women, were trained with a very strong full rich ringing lower range that passed into a refined, powerful, ringing penetrating upper range, but the upper range was not “loud” as we think of loud today, but of course it was loud enough to be heard over ensembles and an entire orchestra. Singers often took flights of fancy, especially sopranos, into the super high notes, but they didn’t stay there, nor did they sing those notes for show. The high note endings we are familiar with are really a more modern invention. As singers began to train more and more as “specific vocal rangers” (dramatic, lyric, coloratura, etc) the placement of the voice changed. Suddenly those high notes we are familiar with were needed, and required, to add excitement to a role. The basic notes of the role, the real tessitura of the role, was lost because such singers could not really give dramatic feeling to anything within the lower part of the staff, and most especially in the chest register as they once did. The dramatic truth of a role was replaced by the ability to “show off” with super loud high notes. And the public went along with it, for whether a singer could hit the high E became more important in the role than whether they could actually do justice to the entire role. Cuts became the norm, and many very dramatic pages of the roles were removed so as to concentrate on the showy.

Unlike today, transposition was acceptable. Sadly, sometimes the transposed version is what we ended up with (example is Lucia di Lammermoor; most of Lucia’s music was transposed DOWN so as to allow the singer to sing a blazing high note ending; the opera flows much better in Donizetti’s original keys, as he was very methodic about how key progression worked in his operas, using key changes to set the dramatic mood; all that is lost with the editions of Lucia we currently use, placing entire acts almost completely in one key).

Even some of the more famous operas are not totally a reflection of what the composer intended, if we know what he intended at all.

Norma is quite different in its original piano score and the decorations we hear in the Sutherland Horne recording in the duets between the two women were part of the original and were published as such. The ending of the act one trio is quite different from what we have today or what we have in the full orchestra version. The closing pages of the trio are completely different only returning to what is familiar to us for the last two pages. What we have now is so completely different. The repeats of the main theme are represented again in the Sutherland Horne recording, but the ending is what we have in our current printed scores. One is left wondering which version was Bellini’s final thoughts on the matter. He actually wrote all the differing versions, and they are found in his manuscript. But he never really indicated which version he wanted, or what extra embellishments in the duets he wanted removed. We currently hear the “Guerra, Guerra” chorus ending with a delicate hynm-like ending that repeats a part of the overture. But that version is NOT in any of the orchestra scores, nor in any of our printed scores commonly available. Yet, it is in the earlier piano rendition and in the manuscript. When was it removed? And by whom?

Even operas by Verdi are not what they seem. Although Verdi did strive to have a “finished final version” of his operas, he didn’t always indicate which final version was his choice. In the French version of Il Trovatore (which is not the Italian version translated into French, but has many textural differences, even completely different music in places, in addition to a ballet) the music has many changes. In the Sutherland Horne Pavarotti recording all the embellishments Sutherland sings will be found in the French version of the opera. Where more difficult lines occur in the Italian version, she simply moves from the French version to the Italian (of course, all sung in Italian). In the Azucena/Manrico duet the mezzo has some really interesting music to sing not in the Italian version, and has a master cadenza taking her to high C and down to low G below middle C. It makes the cadenza in the Italian version (often never sung even as written) look amateurish to the extreme. But it is well in keeping with the mood of the music. The opera actually ends completely differently, with a repeat of the Miserere after the Count comes in to kill Manrico. It is a basic duet with Manrico off-stage and Azucena singing with him. This time Manrico sings what was familiar to us when Leonora sang the Miserere in the first scene of the act, and Azucena reflects those same vocal lines, while the chorus sings something else. The key of the opera is different (A flat instead of E flat that ends the opera in Italian). The closing pronouncements are as we have them in the Italian score, but in the lower key.

Then, we have the multiple versions of Don Carlos. We are all aware of the fact it was written in French in five acts, then in Italian in four acts, then again put into five acts in Italian. But that is hardly a full understanding of that opera. It changes texturally many times in the various versions, with entire parts removed by the time it is in Italian. The Last act is completely different. But it is the part of Eboli that is the interesting challenge. We have it in the score (both in French and Italian) in certain keys, but they are at odds with the role itself. Eboli is written really rather low in most all the music outside the two famous arias (the veil song and the famous O don Fatale). Those arias are placed rather high for the rest of the music of the role. Of course, Verdi must have approved of that, and had a singer who could do it. Now days we have had only a handful of singers who did justice to the role. If they could sing the lower parts effectively, they had to remove the higher notes, or if they could sing the higher notes, the low writing in the ensembles disappears. Even the sad practice of having sopranos singing Eboli has led to hearing two arias sung well, and the rest of the role being passably good, if that good.

Yet, if we go to the score, we discover that Verdi actually wrote the part with various key options. For the most part, the two arias were written and originally published in lower keys. The veil was in G major. O don fatale was originally also in G. In those lower keys they fit the rest of the role much better. And when sung by singers capable of a strong sound in those lower keys, the rest of the role takes on a very noticeable intensity. The tessitura matches. For the veil song, Verdi also gives the aria in 5 different keys, and has 7 different sets of coloratura for the aria. Some of them are more difficult than what is printed, some of them far easier. He is quite emphatic that it is up to the singer to choose which works for them. That is not an option given any singer today.

When Marilyn Horne sang Eboli at the Met she was really panned by the critics for transposing (a thing that happens all the time in Trovatore so the tenor can sing a great high C in the Di Pira, but in reality, he NEVER sings a C, but rather a B or a B flat) these arias. But Verdi gave the options. He approved of the various keys she sang in, and actually provided the different orchestrations in the different keys. She was NOT wrong in using a lower key, and when listening to her version the role actually has much more flare and intensity. Finally we have a singer who is strong in all parts of the role, not just great in two arias.

What is published is not a comfortable sing because the two main arias are actually too high for the rest of the role. The fact we have had singers who had sung what was written effectively doesn’t change the fact that the tessitura is not comfortable.

Also, many roles were written for very specific voices and the writing and role’s tessitura actually is more a reflection of the singers idiosyncrasies than any real specific vocal range.

Semiramide was written by Rossini for his wife, Isabella-Angela Colbran. Although a very gifted singer, she was not without great flaws. But Rossini wrote to reveal her strengths and hide her weakness. Many of Rossini’s most difficult roles were written with her in mind. Yet, they all share one thing in common: they are written on the low side for sopranos, are comfortable for mezzos, and are even possible for contraltos to sing without much transposition. She was one of these singers whose natural voice was contralto, but who was able to manufacture the upper soprano range. She had the rare ability, or rare in that day, of being able to sing upward scales without need of breaking them or having them written in short note groupings. And she could descend equally well. The tessitura of Semiramide is not the usual soprano. It is very low in writing, and even when going to a high B flat, only just gets there and descends. We are told by critics her voice was in ruin by the time she came to sing the role. If that is a ruined voice, I would hate to imagine what her voice in its prime was like. She would have blown us away with her ability.

Today, a good mezzo could easily sing Semiramide.

Below her is the role of Arsace, a contralto. There is no misgivings as to what tessitura this role is. It is definitely a contralto (in fact, one of the lowest roles in all opera for a female voice). Within the written score, the role hardly ventures above the staff at all, while it has much music below.

Tessitura not only helps a role be defined as to who or what voice range should really sing it (as the standard designation “soprano,” “contralto” reveals very little as to the centre of the role) but what area of the singer’s voice will be taxed the most. With many of the roles previously mentioned, the choice is really quite a challenge, for it really does related more to what the various singers singing the roles can endure. Eboli is an extremely low role, excepting two arias. When a singer who can do justice to the majority of the role is required to sing it (like Barbieri when she sang it in London) many of the top notes have to be forgotten, or whole phrases transposed. When a singer with a stunning high C flat sings the role, quite often she cannot sustain the lower singing, especially in ensembles.

Another idiosyncratic role is that of Fides from Meyerbeer’s Le Prophete. The role was expressly written to reflect the talents of Pauline Viardot-Garcia. She was noted as being a mezzo, but one who extended her range up to the soprano high E but also down to the contralto F and G. Her voice was not perfect, nor even even in scale. She had many weaknesses throughout her range, but used those weaknesses as assets in her dramatic presentation. But as one goes through the score of Le Prophete, one sees that the part of Fides is very much, not a standard mezzo role, but definitely a contralto role. The vast majority of the music is set in the middle and lower middle of the staff and very far below the staff. Occasional dramatic outbursts have the singer ascending to high B and high C. There are lines where the singer is required to sing a very strong and loud high B and descend to a G below middle C. The line is seldom a coloratura scale, but a real line of declamatory dramatic singing. But the overall quality of the role, and the tessitura is low. Meyerbeer knew that, and often in places has lower alternatives to those higher lying lines. They are masterfully written, and do not upset the vocal line, nor does one really miss the lost high C which is basically barely touched anyway. Many extremely great contraltos have made a name for themselves with this very difficult role, Ernestine Schumann-Heink being one. In duets and ensembles, Fides is required to carry a vocal line that hardly rises above an A within the staff, but descends below middle C often. Only a true contralto, or mezzo contralto with a good chest register could effectively sing this role (Marilyn Horne is the best example, even though others have sung the notes well, she has the “depth” needed to carry it off).

Now, after all this lengthy discussion of the written note and the history of how things were sung, and the various alterations most roles have, one is left wondering how all this relates to the two arias sung by Urbain in Les Huguenots.

Here we have a character who is actually mixed, or a crossover, of two different tessituras. That is the result of poor editing of the printed score, and a desire to keep all the music Meyerbeer wrote as he wrote it. The most common printed score was edited by Arthur Sullivan. This Royal Edition is also usually in Italian. We will get to why it is the most common score available in a moment. The original French is also available with many noticeable changes from the more common score. Whole sections are added, or actually, retained, in arias that do not appear in the Italian version (particularly in the part of Margherite). And the great aria for Urbain, “Non, non, non, non” is added as an appendix.

As one goes through this score, one clearly sees that Urbain was written for a lyric soprano. There is no mezzo anything about this role. The tessitura is high, but not uncomfortable for that voice range. There are many high Cs, and there are times that Urbain is actually echoing the very vocal line of the Coloratura singing Margherite. The famous “non, non, non” aria is placed within the beginning of act II, and just before a couplet sung with Margherite where both singers sing many high Cs. The aria is definitely that of a contralto. There is nothing about it that is lyric soprano. It descends liberally to low G and even low F all below middle C. Any notes above the staff are few, and for an extremely short duration.

Yet, we find this very contralto music published within the Italian Royal Edition edited by Sullivan. There is sits in the middle of a very lyric soprano role. When we find a singer who does justice to the act I aria (Nobles Seigneurs!) we find a singer who cannot sing the “non, non, non” aria at all, or must sing it in an extremely shortened version with all the lower notes up an octave, thus destroying the flavour and intent of the piece. The evidence of this is that the accompaniment echoes the singer’s descent into the depths, and if the singer does not descend, then the accompaniment doesn’t fit. It was written as a great contralto show piece.

Casting the role is never easy, and for the most part, quite disappointing to listen to. Even if you have a mezzo who can sing all the notes, she often “saves herself” on the high notes singing them quite softly, and does the exact same with the low notes, robbing the role of its personal characteristics.

Why should a role be such a confusion? Well, it is because it was not written as we have it published, or rather it was written in two different ways quite different from the hybrid version we have on paper.

Originally, Urbain was written for a lyric soprano. France had no tradition of mezzos singing the part of youths, boys, or pages. They thought nothing of sopranos singing such roles. The mistake we have today is having most of these soprano roles sung by mezzos, when in point if fact, they were written for sopranos. The tradition of having contraltos singing “male” or travesti roles was an Italian creation. It was a natural outcome of the castrati. When the castrati began to die out, opera composers did NOT replace them with female sopranos, but rather with contraltos. Rossini himself stated that a female contralto is about the closest sound approximation one can get with female voices to the male soprano of a castrato. Female sopranos were almost never used to replace castrati sopranos. And if they were used, it was the Pasta/ Malibran type of singer used, never the Jenny Lind type of nightingale. The castrati were hated in France, and their influence was not felt to the same degree as in Italy.

The best of all representations of this role, and the singing of the first act aria, Nobles Seigneurs, is not any opera singer of today, but rather, Jeannette MacDonald in her film “Maytime.” Not only does she embody the light lyric soprano that was the French ideal, she also sings the aria with the finesse required to make the role sparkling, as it should be. She also does what most all other singers do not do: she sings the opening cadenza on “Seigneurs” in one breath, as it is intended to be sung, with subtle crowning on the important notes that require the singer sit on them just a hair longer. Her actions and her presentation (which do nothing to hide the fact she is a woman – and the French NEVER attempted to hide the fact a woman was singing) are masterful of the period. Her middle cadenza is not what Meyerbeer wrote, but really rather reminiscent of Lucia’s cadenza in her mad scene. Yet, once again, at the end of the aria she delights us with her wondrous breath control as she sings the closing cadenza again in one breath, and still crowning important notes where necessary. Would that modern opera singers do the same. But then, when you have a heavy mezzo or a contralto singing the music, they have to stop to prepare for the physical strength they need to attain the upper notes. That heaviness is not in keeping with the French style.

One can imagine that Jeannette MacDonald would have sung the entire role masterfully (minus the added aria, “non, non, non”).

How did we end up with such a muddled mess of a role and such a mix up in tessitura? Well, that came about because Meyerbeer actually wrote two different versions of the role. For France he wrote what we find mostly in all versions of the score, whether Italian or French. For England he changed many aspects of the part.

England had a tradition of giving all minor soprano roles to mezzos, and all Travesti roles to contraltos, no matter what the original intent of the composer happened to be. For their great presentation of Les Huguenots, they hired the most famous contralto of the day, Marietta Alboni. She was a true contralto, noted for her deep nearly baritonal lower range. But she also had an easy high C. She did sing soprano roles, but always transposed them so as to make them comfortable for her voice. And most all music written for her definitely reflects the tessitura of a real contralto.

For this presentation, Meyerbeer actually prepared her role, and rewrote much of it. The “Nobles Seigneur” was transposed downward to the key of G. The highest note she sang was an A above the staff, and the lowest one an A below the staff. The wondrous echo of Margherite’s coloratura in the second act was altered to something entirely different, with many impressive scales descending to G and F and only once ascending to an A. The textural blending of the two voices (a coloratura and a contralto) is incredible and is far more inventive than the original French version. And for Act II Meyerbeer wrote the famous “Non, non, non” aria. This aria is filled to the brim with descending scales going to many low Gs and one low F. The octave higher options one finds in the score are for the Lyric soprano voice, should that voice sing the aria, but were not included in Alboni’s version. There is hardly a venture above the staff at all. And excepting the G above the staff, all options higher (excepting again the cadenza) are alternates created through editing for the lyric soprano. Only at the cadenza does the singer ascend to a high B flat, which is designed to crown the piece and be a bit of a “show-off” as the entire piece revolves around this page really feeling quite self-important. The modern practice of ending the aria on a high A flat would not have been done by Alboni.

Although the French score was published fairly quickly after the premier, it was the Italian score that became the standard for most opera companies. Originally, the version, as sung by Alboni, was what one found in the published score. Much later the edition we now have was created. The desire of the editors was to combine the music written for France with the best additions created for Covent Garden. And the hybrid we have today was created. No longer did Urbain remain a lyric soprano, but was a mish-mash of lyric soprano and contralto. And ever since this edition was created singers have been forced to combine the very different tessitura.

It is interesting that this Royal Edition was not sung to a French libretto, but to an Italian one. Finding a French score is very hard, but it is available on line. The current score muddies the waters in discovering the true tessitura of the role. That was the result of editing that didn’t take into account the difference in tessitura and just considered that any good singer could sing all the notes. A good singer can, but it takes more than just a good singer to sing the role. It takes the correct singer. If one sings the French version, then a lyric Jeannette MacDonald type singer is needed, and one with as much charm. If the Meyerbeer Italian version is sung, then the singer must be a contralto with a very impressive bottom. And with this version the more modern desire to have the travesti appear to be a boy is acceptable. But for the hybrid score, who would sing it? A contralto would have to transpose to avoid the high notes, so why use one? A lyric would have to ruin the contralto show piece or cut it entirely. The only voice that would do this hybrid well would be another Pasta, Malibran, Viardot-Garcia, a Colbran, or some such singer who had a deep enough voice to make the ‘Non, non, non” aria affective, and the height in the voice to sing all the high Cs throughout the rest of the role. I am sure there are singers who can do it, but based on most modern productions, I would say, casting has not come to terms with the tessitura needed to make the role effective on stage, and musically spectacular to the listener.

We live in a time where transposition is seen as a crime. A very sad thing. That is, if you are not a tenor. It seems tenors can transpose all they want avoiding all sorts of high Cs replacing them with B or B flat. For the rest of us, though, it is seen as almost an insult to the composer.

But transposition made it so many great singers, of not too long ago, were able to really stun us with their singing. Ponselle was noted for transposing the entire first act of Traviata (the first part recitative and Ah forse lui were down a half tone including all the transitional music in F between this part and Sempre Libera; sempre libera was down a full tone). She transposed other numbers within acts, but it was by lowering this first act she was able to sing a most captivating Violetta. It was more moving than imaginable to most viewers. The entire tessitura of the role was suddenly comfortable. Renata Tebaldi did the same, and she also sang a very moving Violetta. Altering that one act, which is written too high in relation to the other acts, placed the tessitura perfect for good singing. Coloraturas have the opposite problem. Many of them will rise the pitch of the other acts, where possible, so as to keep the voice centre comfortable, as the first act is the perfect tessitura for them.

One of the greatest Normas of the twentieth century was Rosa Ponselle. She is still seen as nearly perfect in that role. But she did not sing it as Bellini wrote it. She transposed. We get her transpositions from a copy of a colleague’s score, that of Marion Telva, a contralto. Adalgisa was written for Grisi, a soprano, and the creator of Elvira in I Puritani. She would even herself become a great Norma. But she was a soprano, through and through. Unlike Pasta who was really a mezzo. When the change came to have Adalgisa sung by a contralto or mezzo is not certain, but by the time of the 20th century, it was the norm. Traditionally all the duets between Norma and Adalgisa were transposed downward a tone, and it is in that key that Callas sang them for the most part. Sutherland raised them to score pitch because she had Marilyn Horne who could sing what was written. Even if the contralto was capable of singing the written score, most sang the lower version, as by that time, it was the tradition.

Ponselle sang the written keys for the Casta Diva and the music between it and the cabaletta, “Ah! Bello a me retorna” which she lowered to E. Repeats were not done at that time, or not always, and thus she avoided all high Cs in that part of the score. By the way, the French version of Norma has an aria (very obviously for soprano requiring many High Cs, Ds, and an E) for Adalgisa after her duet with Pollione. I have never heard the aria sung in any production of Norma, but it is a very excellent addition. Of course, she (Ponselle) sang the lower transpositions of the duet between Norma and Adalgisa, again avoiding any high C. The trio that ends the act was dropped a tone, the beginning recitative in B flat (again avoiding the high Cs in the “Oh, non tremare”). The rest of the trio remained in B flat (but there were times the last section “Vanne si” was dropped a tone).

The Act 2 duet was also dropped a tone, the first section in B flat, the “Mira o Norma” in E flat, and the closing allegro also in E flat. At times the scene leading to the “Guerra” chorus was dropped to B flat to avoid the exposed Cs. The Norma Pollione duet was marked “sometimes lowered a tone, sometimes not.” The remainder of the opera remained as printed.

These transpositions, while avoiding high C which may have scared Ponselle, really placed the entire role in a very comfortable tessitura for her to sing. It also made singing the role of Adalgisa wonderfully easy for the contralto required by then to sing that role.

Ponselle was also wise when it came to with whom she sang. She didn’t like singing with bright mezzos because her voice was very dark, like rich red velvet, and she knew that to the public she would sound out of key when singing with a brighter voiced mezzo, for when they sang in thirds, even though she would be singing the higher line, she would sound LOW compared to the mezzo and the public would see her as out of tune. Ponselle had NO pitch problems.

Ponselle used transposition to render the tessitura of a role totally comfortable for her to sing. She even at times transposed the famous “O Patria mia” from Aida down a tone to avoid the High C (which in that opera frightened her greatly). Her desire was to make a role easier to sing, so as to provide her with the ability to perform with confidence.

That is the entire purpose of learning how to evaluate the tessitura of a role in an opera, or a song, so we can sing it with ease and with musicality and confidence. Having all the notes is not nearly as important as having the role fit comfortably on the voice.

What often confuses students as they evaluate a role is they see it with piano only. They learn it with piano only, and they never take the time to see how that comfortable tessitura with piano will feel when combined with orchestra. Piano vocal scores are very deceptive. When full orchestra tutti are written one often sees only octave chords. They do not, no matter how well your pianist plays, produce the same volume or powerful ring of the orchestra. Combining the forces of the brass, the strings, the percussion, etc. even if they are all playing the same note in octaves, is a very overpowering experience. One soon learns that what was easy with piano is nearly impossible with orchestra without resorting to forcing the breath and the tone.

That is an aspect of tessitura that most students don’t consider, and that is not surprising, as they have no understanding of this. I feel that all singers MUST consult a full orchestra score when learning a role. After learning it, take the time to really listen to it on record sung by singers who excelled in that role. Listen to their sound, the power of the voice in those areas where they are required to sing very dramatic music or very emotional utterances. Such music may fall in any part of the voice. When it is in the lower areas of the voice, is the singer still heard? Is the tone full and rich, or is it pressed and pushed? Pressed and pushed we do not want. If so, the tessitura of the role was wrong for the voice. Even if he/she could sing the notes, he/she could not sing them correctly. Does the voice seem to nearly fall apart during pianissimi in the upper range? If so, then reconsider how the voice was used in more dramatic music in the middle and lower range. Misuse of those ranges robs the upper range of stability.

It isn’t just the notes and where they sit most in a role that defines it (tessitura), but also what is required of the voice in all parts of the range, and especially in that area most used. Even if all the notes are comfortable for regular singing, if they become a stress when put to the test in the drama and with the orchestra, then the tessitura is wrong for your voice.

It is these factors that convinced Ebe Stagnini not to sing Norma on stage, even though at the piano it was quite easily sung. She knew the requirements of the role. She knew what it required in vocal heft when combined with orchestra and chorus. Amneris may have been perfect for her, and requiring just as much heft (and she was a singer with a truly huge ringing voice, nothing could overpower her) but Amneris is not written like Norma. Its centre is quite different. The requirements very different. She was able to meet them easily.

The purpose of this very long essay is to help singers understand what tessitura really is, and how it plays a huge part in what we choose to sing. And that it is much more than simply knowing where the central notes are and knowing they are well within your range. It is knowing what is required at all times and in all areas of the voice to create a real living character, and to make the music alive to the listener. And it is to do all that without stressing or pushing the voice. Doing it with relative ease so as to have full confidence in what you are doing. Nothing is worse than singing a role where you are on pins and needles all the time “hoping” you will endure it enough to finish. That should be no singer’s goal. Understanding tessitura is what we must do in order to learn to sing and choose roles most perfect for our individual voices. Only then, can we excel to our very best in all we do.